Abstract
Childhood apraxia of speech (CAS) commonly affects the production of lexical stress contrast in polysyllabic words. Automated classification tools have the potential to increase reliability and efficiency in measuring lexical stress. Here, factors affecting the accuracy of a custom-built deep neural network (DNN)-based classification tool are evaluated. Sixteen children with typical development (TD) and 26 with CAS produced 50 polysyllabic words. Words with strong–weak (SW, e.g., dinosaur) or WS (e.g., banana) stress were fed to the classification tool, and the accuracy measured (a) against expert judgment, (b) for speaker group, and (c) with/without prior knowledge of phonemic errors in the sample. The influence of segmental features and participant factors on tool accuracy was analysed. Linear mixed modelling showed significant interaction between group and stress type, surviving adjustment for age and CAS severity. For TD, agreement for SW and WS words was >80%, but CAS speech was higher for SW (>80%) than WS (~60%). Prior knowledge of segmental errors conferred no clear advantage. Automatic lexical stress classification shows promise for identifying errors in children’s speech at diagnosis or with treatment-related change, but accuracy for WS words in apraxic speech needs improvement. Further training of algorithms using larger sets of labelled data containing impaired speech and WS words may increase accuracy.
Highlights
Difficulty with the production of lexical stress has been identified as one of the core deficits in both childhood and acquired apraxia of speech (CAS and AOS, respectively) [1,2] and has been studied for its potential as a diagnostic marker (e.g., [3,4,5])
This study aims to further the work of Shahin and colleagues [9,10,11,12] in the development of an automated lexical stress classification tool for Childhood apraxia of speech (CAS)
We compare a tool-based classification of stress patterns in isolated polysyllabic words (PSW) with expert auditory perceptual judgment given that lexical stress in PSW has proven sensitive to the differential diagnosis of CAS (e.g., [4])
Summary
Difficulty with the production of lexical stress has been identified as one of the core deficits in both childhood and acquired apraxia of speech (CAS and AOS, respectively) [1,2] and has been studied for its potential as a diagnostic marker (e.g., [3,4,5]). Assessment of lexical stress production is traditionally impressionistic [6] and vulnerable to various sources of error and bias within and between rater [7]. This study aims to further the work of Shahin and colleagues [9,10,11,12] in the development of an automated lexical stress classification tool for CAS. We compare a tool-based classification of stress patterns in isolated polysyllabic words (PSW) with expert auditory perceptual judgment given that lexical stress in PSW has proven sensitive to the differential diagnosis of CAS (e.g., [4]). We explore the potential for knowledge-driven systems to boost tool-based classification accuracy for mispronounced words and examine classification errors for potential segmental factors, which may affect tool accuracy and so guide stimulus selection for reliable assessment instruments in the future
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
